MAI Image 2.7: Microsoft cuts the cost of AI image generation and editing
Microsoft has launched two new image models, MAI Image 2.7 and MAI Image 2.7 Flash, along with a creative workspace called Canvas. The company is promising more precise edits and substantially lower pricing than several premium image models.
Microsoft is rapidly expanding its in-house model portfolio. On November 17, the company introduced MAI Image 2.7, MAI Image 2.7 Flash and Canvas, a new interactive environment for creating and editing images. Both models are available in public preview through Microsoft Foundry.
MAI Image 2.7 is positioned as the higher-quality option. Microsoft says it can change an object’s colour, material or position without unnecessarily altering the rest of the image. That matters to creative teams because a generative edit can easily damage details that were supposed to remain unchanged, including a background, a product or the overall composition.
The Flash version is built for speed and high-volume production. It also powers Canvas, where users can generate an image, request variations and make successive edits within the same workspace. Microsoft is therefore targeting two different needs: projects where quality and fine control matter most, and repetitive jobs where turnaround time and cost are the priority.
Pricing is a central part of the announcement. Microsoft lists an approximate cost of US$35 per 1,000 images for MAI Image 2.7 and about US$10 per 1,000 images for Flash. Those prices could appeal to organizations producing catalogues, advertising concepts, product variations or large amounts of digital content. They do not necessarily include every related expense, such as integration, storage, human review and governance.
Microsoft also points to favourable results in third-party evaluations. As of November 16, MAI Image 2.7 ranked first on Artificial Analysis’ AA-Image-Editing v2.1 leaderboard, with an Elo score of 1,215 based on 5,478 votes. Its lead over the next model was only four points, however. That narrow difference is better interpreted as evidence that the model is competitive, rather than proof that it will be the best choice for every project.
The model card published by Microsoft describes MAI Image 2.7 as a diffusion model designed for both text-to-image generation and controlled image-to-image editing. The document provides information about the developer and the model’s release, but it does not replace an independent review of its limitations. Teams should still evaluate visual consistency, fine-detail reproduction, copyright management and the mechanisms used to identify synthetic content.
For creators, the most useful aspect may be the combination of a faster model and a more precise one. A team could use Flash to explore many visual directions, then switch to MAI Image 2.7 for final refinements. That workflow may reduce costs, provided human review remains part of the process and the origin of visual assets is properly documented.
MAI Image 2.7 also reinforces Microsoft’s larger strategy: offering specialized first-party models inside Foundry instead of relying exclusively on outside developers. Its real quality will have to be confirmed in production, particularly when a brand’s identity or an accurate representation of a product is essential.
Key points
- MAI Image 2.7 focuses on image quality and precise editing control.
- MAI Image 2.7 Flash prioritizes speed and high-volume workloads.
- Listed prices are approximately US$35 and US$10 per 1,000 images.
- Canvas provides an interactive environment for image creation and editing.
- The Artificial Analysis ranking is promising, but the lead over the next model is narrow.
Sources
- Microsoft AI, “MAI Image 2.7: Frontier quality at a fraction of the cost,” November 17, 2026: https://microsoft.ai/news/introducing-mai-image-2-7/
- Microsoft AI, “Seven new world class models at scalable price-performance,” November 17, 2026: https://microsoft.ai/news/microsoft-ignite-2026/
- Microsoft, “MAI Image 2.7 / MAI Image 2.7 Flash Model Card,” released November 17, 2026: https://microsoft.ai/pdf/MAI-Image-2.7-Model-Card.pdf



